Probabilistic modeling of dam failure scenarios: a case study of Kanlikoy Dam in Cyprus

Abstract

One of the most perilous natural hazards is flooding resulting from dam failure, which can devastate downstream infrastructure and lead to significant human casualties. In recent years, the frequency of flash floods in the northern part of Nicosia, Cyprus, has increased. This area faces increased risk as it lies downstream of the Kanlikoy Dam, an aging earth-fill dam constructed over 70 years ago. In this study, we aim to assess potential flood hazards stemming from three distinct failure scenarios: piping, 100-year rainfall, and probable maximum precipitation (PMP). To achieve this, HEC-HMS hydrologic model findings were integrated into 2D HEC-RAS hydraulic models to simulate flood hydrographs and generate flood inundation and hazard maps. For each scenario, Monte Carlo simulations using McBreach software produced four hydrographs corresponding to exceedance probabilities of 90%, 50%, 10%, and 1%. The results indicate that all dam breach scenarios pose a significant threat to agricultural and residential areas, leading to the destruction of numerous buildings, roads, and infrastructures. Particularly, Scenario 3, which includes PMP, was identified as the most destructive, resulting in prevailing flood hazard levels of H5 and H6 in the inundated areas. The proportion of inundated areas in these high hazard levels varied between 52.8% and 57.4%, with the number of vulnerable structures increasing from 248 to 321 for exceedance probabilities of 90% and 1%, respectively. Additionally, the number of flooded buildings ranged from 842 to 935, and 26 to 34 km of roads were found to be inundated in this scenario. These findings revealed the need for authorities to develop comprehensive evacuation plans and establish an efficient warning system to mitigate the flood risks associated with dam failure.